Computer science education, at least at the introductory level, was once monolithic. The first programming course at most schools would cover pretty much the same material, from the same language--although that language did change every ten years or so. This is clearly no longer true, as the field has broadened and split into a number of subdisciplines, including the following from the Association for Computing Machinery (ACM) Computing Curricula 2005: computer engineering, computer science (CS), information systems (IS), information technology (IT), and software engineering. In a worst-case scenario, each of these subdisciplines, as well as others such as multimedia, could have its own introductory course, its own second programming course, and so on. The potential for fragmentation and proliferation, not to mention student confusion, could easily turn into a nightmare.

Bills and Canosa come from two departments at the Rochester Institute of Technology (RIT): IT and CS. RIT is fortunate to have a college of computing and information sciences, with a number of ABET-accredited programs. In fact, in addition to IT and CS, there are two other separate departments, one for software engineering and one for network security and systems administration. Therefore, the problems of fragmentation and proliferation are even more acute at RIT than at most schools.

The paper starts with an excellent, unbiased history of the evolution of the computing discipline, with discussion of the major attempts by national organizations. These organizations, including the ACM, the Institute of Electrical and Electronics Engineers Computer Society (IEEE-CS), the Association of Information Technology Professionals (AITP), and the Association for Information Systems (AIS), provide curriculum guidance, with an emphasis on the ACM Computing Curricula 2005.

Bills and Canosa then perform three very valuable services: they identify programming as a foundation skill for all aspects of computer science and state the need to find some underlying programming topics that are common to all of the subdisciplines; they recognize that the list of common subject matter will by necessity not be complete--each field will be forced to add its own topics; and they propose a mechanism for constructing a first-year curriculum in which students take a series of modules, some of which would be common and some of which would be specific to a subdiscipline.

The sample combined first-year curriculum geared toward Bills and Canosa’s departments at RIT. They do, of course, recognize that the decision on how to divide topics must be done at the local level, based on the characteristics of the individual institution, including “both politics and assessment issues.” However, we can all hope that their final conclusion, that “the flexible core and depth approach outlined here has potential for success,” proves to be accurate. Even if this prediction is too sanguine, the paper’s attempt to encourage curriculum sharing makes a significant contribution to the field.